Re: [PATCH] bootstrap: support --no-git option

On 01/20/2011 11:54 AM, Eric Blake wrote: > @@ -64,6 +64,9 @@ Options: > --force Attempt to bootstrap even if the sources seem >not to have been checked out. > --skip-poDo not download po files. > + --no-git Do not upd

mbiter, mbfile: cleanup old cruft

This patch completes the one from 2008-12-21 . 2011-01-20 Bruno Haible mbfile, mbiter: Complete change from 2008-12-21. * m4/mbfile.m4 (gl_MBFILE): Don't require AC_FUNC_MBRTOWC. * m4/mbiter.m4 (gl_MBI

Re: speed up u8_strstr

On 01/20/2011 01:16 PM, Pádraig Brady wrote: > On 31/07/10 20:48, Bruno Haible wrote: >> Hi Pádraig, >> the u8_strstr function. >>> >>> I wonder could we speed that up for UTF-8 >>> by just deferring to strstr() ? >> >> This is a good idea, because the naïve implementation of u8_strstr is >> q

Re: speed up u8_strstr

On 31/07/10 20:48, Bruno Haible wrote: > Hi Pádraig, > >>> the u8_strstr function. >> >> I wonder could we speed that up for UTF-8 >> by just deferring to strstr() ? > > This is a good idea, because the naïve implementation of u8_strstr is > quadratic (O(n²)), whereas for strstr we now have an O(

[PATCH] bootstrap: support --no-git option

From: Benjamin Lindner * build-aux/bootstrap: Add --no-git option, to be used when --gnulib-srcdir points to the exact desired checkout. Signed-off-by: Eric Blake --- I'd like to push this in your name, once I have your okay. The (tiny change) notation reflects your copyright status, and is n

Re: gnulib unnecessarily enforces git as requirement

On 01/20/2011 09:50 AM, Benjamin Lindner wrote: >> >> where the new --no-git option is what makes it explicit that you intend >> to use an existing directory as-is rather than update a submodule in >> your local tree. > > Ok, first shot, see attached patch. > > With this, > this works:export

Re: gnulib unnecessarily enforces git as requirement

> Absolutely. > > One idea would be: > > --gnulib-srcdir overrides environment, but defaults to $GNULIB_SRCDIR > > bootstrap                            require git, creates fresh clone > bootstrap --gnulib-srcdir            require git, clones using srcdir as > reference > bootstrap --no-git      

Re: gnulib unnecessarily enforces git as requirement

On 01/20/2011 09:08 AM, Benjamin Lindner wrote: >> git IS used by the bootstrap script except for the one case where you >> specifically ask that it not be used. So, the best thing might be to >> enhance the bootstrap pre-requisite checks to see if --gnulib-srcdir is >> in effect and use that as a

Re: gnulib unnecessarily enforces git as requirement

> git IS used by the bootstrap script except for the one case where you > specifically ask that it not be used.  So, the best thing might be to > enhance the bootstrap pre-requisite checks to see if --gnulib-srcdir is > in effect and use that as a reason to skip the git prereq check. That sounds s

Re: gnulib unnecessarily enforces git as requirement

Benjamin Lindner wrote: ... > And - closing the circle to the start of the report - since actually > git not required anyway, why add the extra effort of msysgit? Eric Blake explained in detail: http://thread.gmane.org/gmane.comp.lib.gnulib.bugs/24825/focus=24828 perhaps you haven't seen his r

Re: gnulib unnecessarily enforces git as requirement

>> I am using msys as build environment and git is not available there. > > From what I recall, git has been available on msys for some time. > Have you looked at this? > >  http://code.google.com/p/msysgit/ Yes I have, but they use a tweaked (patched) version of msys, so it follows that it is not

Re: [PATCH] {maint} tests: new subroutines for test skipping/failing

Ralf Wildenhues wrote: ... >> > > +# This is useful when using automake's parallel tests mode, to print >> > > +# the reason for skip/failure to console, rather than to the .log files. >> > > +: ${stderr_fileno_=2} >> > > + >> > > +warn_() { echo "$@" 1>&$stderr_fileno_; } >> > > +fail_() { warn_ "

Re: gnulib unnecessarily enforces git as requirement

Benjamin Lindner wrote: > Octave (http://www.octave.org) utilizes gnulib and a recent change in > gnulib breaks the build for MinGW platform, because gnulib now > enforces git to be available at bootstrap stage. > The change was introduced in a patch discussed in the following thread: > http://list

Re: gnulib unnecessarily enforces git as requirement

On 01/20/2011 06:26 AM, Benjamin Lindner wrote: > Hello list, > > Octave (http://www.octave.org) utilizes gnulib and a recent change in > gnulib breaks the build for MinGW platform, because gnulib now > enforces git to be available at bootstrap stage. > The change was introduced in a patch discuss

Re: Interix list of mounted file systems.

Markus Duft wrote: > On 01/20/2011 09:53 AM, Jim Meyering wrote: > [snip] >> >> There was no need to rebase. >> However, there was no ChangeLog entry, and the style of your added >> function did not match the existing code (we require a space before >> most open parentheses, and declare pointers li

Re: [PATCH] maint.mk: improve the public-submodule-commit rule

On 01/20/2011 04:34 AM, Jim Meyering wrote: > This is in preparation for hooking this rule to "check" > at least in coreutils. Today I was burned once again by > pushing a non-public submodule reference (coreutils). > The fact that this check is currently run only at release time > is good, but ob

gnulib unnecessarily enforces git as requirement

Hello list, Octave (http://www.octave.org) utilizes gnulib and a recent change in gnulib breaks the build for MinGW platform, because gnulib now enforces git to be available at bootstrap stage. The change was introduced in a patch discussed in the following thread: http://lists.gnu.org/archive/htm

Re: Interix list of mounted file systems.

On 01/20/2011 09:53 AM, Jim Meyering wrote: [snip] > > There was no need to rebase. > However, there was no ChangeLog entry, and the style of your added > function did not match the existing code (we require a space before > most open parentheses, and declare pointers like char *p, not "char* p"),

[PATCH] maint.mk: improve the public-submodule-commit rule

This is in preparation for hooking this rule to "check" at least in coreutils. Today I was burned once again by pushing a non-public submodule reference (coreutils). The fact that this check is currently run only at release time is good, but obviously not enough. It probably belongs on a server-s

Re: bug#7859: syncing Emacs from sources maintained elsewhere

Paul Eggert skrev 2011-01-20 08:51: On 01/19/2011 10:45 PM, Jan Djärv wrote: /* config.h #define:s malloc/realloc/free and then includes stdlib.h. We want the undefined versions, but if config.h includes stdlib.h with the #define:s in place, the prototypes will be wrong and we get wa

Re: [PATCH] include_next: do not check for standard headers like stddef.h

Hi Paul, > * m4/include_next.m4 (gl_NEXT_HEADERS): New macro, which does not > use AC_CHECK_HEADERS_ONCE, but which otherwise contains what > gl_CHECK_NEXT_HEADERS used to contain. This makes 'configure' run > faster for headers like stddef.h that are known to exist. While doing this, you overlo

Re: Suppressing GNULIB_TEST_MKTIME etc. in emacs/src/config.in

Hi Paul, > put the following into configure.in: > >AC_DEFUN([gl_MODULE_INDICATOR_FOR_TESTS], []) > > This zaps the macro that generates the unwanted GNULIB_TEST_ indicators. This is a good enough solution, IMO. > * Define a new macro in gnulib that does the above AC_DEFUN, then >invok

Re: [PATCH] gl_CHECK_NEXT_HEADERS implies AC_CHECK_HEADERS_ONCE

Hi Paul, > Here's a proposed patch to simplify gnulib's .m4 files slightly. > It doesn't affect behavior; it's just a minor refactoring. So far these .m4 files use AC_CHECK_HEADERS_ONCE explicitly because the fact that gl_CHECK_NEXT_HEADERS invokes AC_CHECK_HEADERS_ONCE is an undocumented impleme

Re: Interix list of mounted file systems.

Markus Duft wrote: > On 10/21/2010 12:35 PM, Jim Meyering wrote: >> Markus Duft wrote: >>> On 10/21/2010 11:59 AM, James Youngman wrote: On Thu, Oct 21, 2010 at 10:05 AM, Markus Duft wrote: > Hi! > > I just created the attached patch for findutils to build. Any chance to > get

Suppressing GNULIB_TEST_MKTIME etc. in emacs/src/config.in

Now that Emacs is using a bit of gnulib, its src/config.in contains three #undefs that look like this: /* Define to 1 when the gnulib module mktime should be tested. */ #undef GNULIB_TEST_MKTIME These symbols are unused by Emacs: it doesn't have gnulib tests and is not likely to have them s